The Department of Electrical and Computer Engineering faculty member Dr. Vincent Tan has been awarded the National Research Foundation (NRF) Fellowship (Class of 2018).
The fellowship, which carries a research grant of up to $2.5 million, provides opportunities for early career researchers to carry out independent research locally over a five-year period. The title of his project is “Fundamental Performance Limits for Statistical Learning Algorithms”.
Dr. Tan’s main research program aims to answer the following question that is salient today’s data-rich environment: Can one systematically derive fundamental achievability and impossibility results for the performances of various statistical machine learning tasks?
The overarching goal in data science is to develop algorithms that are computationally efficient and can generalize well. However, while many of the state-of-the-art algorithms perform reasonably well in practice, one aspect that researchers have largely neglected is the derivation of fundamental performance limits, without regard to computational complexity considerations. While the speed of light is not attainable for the design of cars and airplanes, the main thrust in this research program is to show that the performances of some statistical learning tasks can indeed come close to how well they can ultimately perform.
The techniques used will primarily be information-theoretic (IT) in nature. This allows us to establish of achievability and impossibility results in the absence of computational complexity requirements and to compare the performance of the state-of-the-art algorithms to their newly-discovered fundamental performance limits.
The NRF funding will also allow him to develop a world-class team and to make progress on developing fundamental limits from problems ranging from privacy-preserving learning to dictionary learning (matrix factorization) and ranking.
Dr. Vincent Tan is supported by the National Research Foundation (NRF), Prime Minister’s Office, Singapore, under its NRF Fellowship (Class of 2018).
